End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- id
 license: mit
 base_model: indolem/indobert-base-uncased
 tags:

all_results.json

@@ -1,21 +1,21 @@
 {
-    "accuracy": 0.9173306772908366,
+    "accuracy": 0.9183266932270916,
     "epoch": 20.0,
-    "eval_accuracy": 0.8847117794486216,
-    "eval_f1": 0.8609292598654301,
-    "eval_loss": 0.2950780689716339,
-    "eval_precision": 0.8609292598654301,
-    "eval_recall": 0.8609292598654301,
-    "eval_runtime": 4.9799,
+    "eval_accuracy": 0.8822055137844611,
+    "eval_f1": 0.855319904024935,
+    "eval_loss": 0.3026413023471832,
+    "eval_precision": 0.862378106322743,
+    "eval_recall": 0.8491543917075832,
+    "eval_runtime": 1.8078,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.122,
-    "eval_steps_per_second": 10.04,
-    "f1": 0.8996936595347718,
-    "precision": 0.9017567207878627,
-    "recall": 0.8977052792948165,
-    "train_loss": 0.2363556080177182,
-    "train_runtime": 1909.167,
+    "eval_samples_per_second": 220.713,
+    "eval_steps_per_second": 27.658,
+    "f1": 0.9005988602337487,
+    "precision": 0.9039776379609248,
+    "recall": 0.897413049726933,
+    "train_loss": 0.24652244458433056,
+    "train_runtime": 623.7969,
     "train_samples": 3645,
-    "train_samples_per_second": 38.184,
-    "train_steps_per_second": 1.278
+    "train_samples_per_second": 116.865,
+    "train_steps_per_second": 3.912
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.8847117794486216,
-    "eval_f1": 0.8609292598654301,
-    "eval_loss": 0.2950780689716339,
-    "eval_precision": 0.8609292598654301,
-    "eval_recall": 0.8609292598654301,
-    "eval_runtime": 4.9799,
+    "eval_accuracy": 0.8822055137844611,
+    "eval_f1": 0.855319904024935,
+    "eval_loss": 0.3026413023471832,
+    "eval_precision": 0.862378106322743,
+    "eval_recall": 0.8491543917075832,
+    "eval_runtime": 1.8078,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.122,
-    "eval_steps_per_second": 10.04
+    "eval_samples_per_second": 220.713,
+    "eval_steps_per_second": 27.658
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.9173306772908366,
-    "f1": 0.8996936595347718,
-    "precision": 0.9017567207878627,
-    "recall": 0.8977052792948165
+    "accuracy": 0.9183266932270916,
+    "f1": 0.9005988602337487,
+    "precision": 0.9039776379609248,
+    "recall": 0.897413049726933
 }

predict_results.txt

@@ -13,8 +13,8 @@ index	prediction
 11	1
 12	0
 13	1
-14	1
-15	1
+14	0
+15	0
 16	0
 17	1
 18	1
@@ -37,16 +37,16 @@ index	prediction
 35	1
 36	1
 37	1
-38	1
+38	0
 39	0
 40	0
 41	1
-42	1
+42	0
 43	1
 44	1
 45	1
 46	1
-47	1
+47	0
 48	0
 49	1
 50	1
@@ -57,7 +57,7 @@ index	prediction
 55	1
 56	1
 57	1
-58	0
+58	1
 59	1
 60	1
 61	1
@@ -85,7 +85,7 @@ index	prediction
 83	1
 84	1
 85	1
-86	0
+86	1
 87	1
 88	1
 89	1
@@ -113,14 +113,14 @@ index	prediction
 111	1
 112	1
 113	1
-114	1
+114	0
 115	1
 116	1
 117	1
 118	1
 119	1
 120	1
-121	0
+121	1
 122	1
 123	1
 124	1
@@ -152,11 +152,11 @@ index	prediction
 150	1
 151	1
 152	1
-153	1
+153	0
 154	1
 155	1
 156	1
-157	1
+157	0
 158	1
 159	1
 160	1
@@ -192,7 +192,7 @@ index	prediction
 190	0
 191	1
 192	1
-193	0
+193	1
 194	1
 195	1
 196	1
@@ -207,7 +207,7 @@ index	prediction
 205	1
 206	1
 207	0
-208	1
+208	0
 209	1
 210	1
 211	1
@@ -220,11 +220,11 @@ index	prediction
 218	1
 219	1
 220	1
-221	1
+221	0
 222	1
 223	1
 224	1
-225	0
+225	1
 226	1
 227	1
 228	1
@@ -245,7 +245,7 @@ index	prediction
 243	1
 244	1
 245	1
-246	0
+246	1
 247	1
 248	1
 249	1
@@ -258,7 +258,7 @@ index	prediction
 256	1
 257	1
 258	1
-259	0
+259	1
 260	0
 261	1
 262	1
@@ -271,7 +271,7 @@ index	prediction
 269	0
 270	1
 271	1
-272	0
+272	1
 273	1
 274	1
 275	1
@@ -288,7 +288,7 @@ index	prediction
 286	1
 287	0
 288	1
-289	0
+289	1
 290	1
 291	0
 292	1
@@ -300,7 +300,7 @@ index	prediction
 298	0
 299	0
 300	0
-301	1
+301	0
 302	0
 303	0
 304	0
@@ -356,10 +356,10 @@ index	prediction
 354	0
 355	0
 356	0
-357	0
+357	1
 358	0
 359	0
-360	1
+360	0
 361	0
 362	0
 363	0
@@ -396,7 +396,7 @@ index	prediction
 394	0
 395	0
 396	1
-397	1
+397	0
 398	0
 399	0
 400	0
@@ -423,7 +423,7 @@ index	prediction
 421	0
 422	0
 423	0
-424	0
+424	1
 425	0
 426	0
 427	0
@@ -436,7 +436,7 @@ index	prediction
 434	0
 435	0
 436	0
-437	1
+437	0
 438	0
 439	0
 440	0
@@ -536,7 +536,7 @@ index	prediction
 534	0
 535	0
 536	0
-537	0
+537	1
 538	0
 539	0
 540	0
@@ -558,7 +558,7 @@ index	prediction
 556	0
 557	0
 558	0
-559	1
+559	0
 560	0
 561	0
 562	0
@@ -610,7 +610,7 @@ index	prediction
 608	0
 609	0
 610	0
-611	0
+611	1
 612	0
 613	0
 614	0
@@ -626,7 +626,7 @@ index	prediction
 624	0
 625	0
 626	0
-627	1
+627	0
 628	0
 629	0
 630	0
@@ -656,7 +656,7 @@ index	prediction
 654	0
 655	0
 656	0
-657	1
+657	0
 658	1
 659	0
 660	0
@@ -716,7 +716,7 @@ index	prediction
 714	0
 715	0
 716	0
-717	1
+717	0
 718	0
 719	0
 720	0
@@ -753,7 +753,7 @@ index	prediction
 751	0
 752	0
 753	0
-754	1
+754	0
 755	0
 756	0
 757	0
@@ -762,7 +762,7 @@ index	prediction
 760	0
 761	0
 762	0
-763	0
+763	1
 764	0
 765	0
 766	0
@@ -772,13 +772,13 @@ index	prediction
 770	0
 771	0
 772	0
-773	1
-774	0
+773	0
+774	1
 775	0
 776	0
 777	0
 778	0
-779	1
+779	0
 780	0
 781	0
 782	0
@@ -792,11 +792,11 @@ index	prediction
 790	0
 791	0
 792	0
-793	0
+793	1
 794	0
 795	0
 796	0
-797	0
+797	1
 798	1
 799	0
 800	0
@@ -851,7 +851,7 @@ index	prediction
 849	0
 850	0
 851	0
-852	0
+852	1
 853	0
 854	0
 855	0

runs/Jun03_15-01-46_a358b85c7679/events.out.tfevents.1717427545.a358b85c7679.158543.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9106848a02fa4c3a2a8c888b728e37234caabffdf8b01b4ebb53776969a294fa
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.2363556080177182,
-    "train_runtime": 1909.167,
+    "train_loss": 0.24652244458433056,
+    "train_runtime": 623.7969,
     "train_samples": 3645,
-    "train_samples_per_second": 38.184,
-    "train_steps_per_second": 1.278
+    "train_samples_per_second": 116.865,
+    "train_steps_per_second": 3.912
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 2.953822374343872,
+      "grad_norm": 3.928847551345825,
       "learning_rate": 4.75e-05,
-      "loss": 0.5448,
+      "loss": 0.5514,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7243107769423559,
-      "eval_f1": 0.6567661411411412,
-      "eval_loss": 0.5046952366828918,
-      "eval_precision": 0.6629251700680272,
-      "eval_recall": 0.6524368066921259,
-      "eval_runtime": 5.0686,
-      "eval_samples_per_second": 78.72,
-      "eval_steps_per_second": 9.865,
+      "eval_accuracy": 0.7218045112781954,
+      "eval_f1": 0.6545993371027491,
+      "eval_loss": 0.5084339380264282,
+      "eval_precision": 0.6600553802562947,
+      "eval_recall": 0.6506637570467357,
+      "eval_runtime": 1.7877,
+      "eval_samples_per_second": 223.186,
+      "eval_steps_per_second": 27.968,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 2.6171164512634277,
+      "grad_norm": 3.589020013809204,
       "learning_rate": 4.5e-05,
-      "loss": 0.4527,
+      "loss": 0.4753,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7944862155388471,
-      "eval_f1": 0.775152564736929,
-      "eval_loss": 0.4319760501384735,
-      "eval_precision": 0.7667055167055168,
-      "eval_recall": 0.8121022004000726,
-      "eval_runtime": 5.0454,
-      "eval_samples_per_second": 79.082,
-      "eval_steps_per_second": 9.91,
+      "eval_accuracy": 0.8170426065162907,
+      "eval_f1": 0.7662453352594198,
+      "eval_loss": 0.40073099732398987,
+      "eval_precision": 0.7870255775577557,
+      "eval_recall": 0.7530460083651573,
+      "eval_runtime": 1.8277,
+      "eval_samples_per_second": 218.302,
+      "eval_steps_per_second": 27.356,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 4.439111709594727,
+      "grad_norm": 7.217240810394287,
       "learning_rate": 4.25e-05,
-      "loss": 0.3603,
+      "loss": 0.3834,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8471177944862155,
-      "eval_f1": 0.7984931903796002,
-      "eval_loss": 0.33696243166923523,
-      "eval_precision": 0.8393082695594026,
-      "eval_recall": 0.7768230587379523,
-      "eval_runtime": 5.0496,
-      "eval_samples_per_second": 79.016,
-      "eval_steps_per_second": 9.902,
+      "eval_accuracy": 0.8395989974937343,
+      "eval_f1": 0.7805280528052805,
+      "eval_loss": 0.35415172576904297,
+      "eval_precision": 0.8448765432098766,
+      "eval_recall": 0.7540007274049827,
+      "eval_runtime": 1.8275,
+      "eval_samples_per_second": 218.333,
+      "eval_steps_per_second": 27.36,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 0.9325271248817444,
+      "grad_norm": 1.0342143774032593,
       "learning_rate": 4e-05,
-      "loss": 0.3081,
+      "loss": 0.3188,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8461962888779714,
-      "eval_loss": 0.29947736859321594,
-      "eval_precision": 0.8453465227094517,
-      "eval_recall": 0.8470631023822512,
-      "eval_runtime": 5.0527,
-      "eval_samples_per_second": 78.967,
-      "eval_steps_per_second": 9.896,
+      "eval_accuracy": 0.8621553884711779,
+      "eval_f1": 0.8333016825553572,
+      "eval_loss": 0.32140952348709106,
+      "eval_precision": 0.8341507249908615,
+      "eval_recall": 0.8324695399163484,
+      "eval_runtime": 1.8265,
+      "eval_samples_per_second": 218.445,
+      "eval_steps_per_second": 27.374,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 1.1945856809616089,
+      "grad_norm": 1.17711341381073,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2793,
+      "loss": 0.2981,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.87468671679198,
-      "eval_f1": 0.8456742372671576,
-      "eval_loss": 0.3008015751838684,
-      "eval_precision": 0.8536697247706422,
-      "eval_recall": 0.8388343335151845,
-      "eval_runtime": 5.0466,
-      "eval_samples_per_second": 79.063,
-      "eval_steps_per_second": 9.908,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.855319904024935,
+      "eval_loss": 0.2984267771244049,
+      "eval_precision": 0.862378106322743,
+      "eval_recall": 0.8491543917075832,
+      "eval_runtime": 1.8288,
+      "eval_samples_per_second": 218.18,
+      "eval_steps_per_second": 27.341,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 3.8788158893585205,
+      "grad_norm": 4.153679370880127,
       "learning_rate": 3.5e-05,
-      "loss": 0.2526,
+      "loss": 0.2835,
       "step": 732
     },
     {
       "epoch": 6.0,
       "eval_accuracy": 0.8696741854636592,
-      "eval_f1": 0.8411818738518064,
-      "eval_loss": 0.2986568510532379,
-      "eval_precision": 0.8448542607834644,
-      "eval_recall": 0.8377886888525186,
-      "eval_runtime": 5.0571,
-      "eval_samples_per_second": 78.899,
-      "eval_steps_per_second": 9.887,
+      "eval_f1": 0.8368354828562441,
+      "eval_loss": 0.28104665875434875,
+      "eval_precision": 0.8520237470480189,
+      "eval_recall": 0.8252864157119476,
+      "eval_runtime": 1.8275,
+      "eval_samples_per_second": 218.335,
+      "eval_steps_per_second": 27.36,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 1.2110929489135742,
+      "grad_norm": 1.0657362937927246,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.2478,
+      "loss": 0.2517,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.8771929824561403,
-      "eval_f1": 0.846679266293906,
-      "eval_loss": 0.3030013144016266,
-      "eval_precision": 0.8609191655801824,
-      "eval_recall": 0.8356064739043463,
-      "eval_runtime": 5.0491,
-      "eval_samples_per_second": 79.023,
-      "eval_steps_per_second": 9.903,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8622036668943447,
+      "eval_loss": 0.28660351037979126,
+      "eval_precision": 0.8671602787456446,
+      "eval_recall": 0.8577014002545917,
+      "eval_runtime": 1.7903,
+      "eval_samples_per_second": 222.862,
+      "eval_steps_per_second": 27.928,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 3.5065371990203857,
+      "grad_norm": 3.21960711479187,
       "learning_rate": 3e-05,
-      "loss": 0.2337,
+      "loss": 0.2374,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8671679197994987,
-      "eval_f1": 0.8350789627607721,
-      "eval_loss": 0.2974022924900055,
-      "eval_precision": 0.8463358876939919,
-      "eval_recall": 0.8260138206946717,
-      "eval_runtime": 5.0637,
-      "eval_samples_per_second": 78.796,
-      "eval_steps_per_second": 9.874,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8485289465359063,
+      "eval_loss": 0.2996794879436493,
+      "eval_precision": 0.8671008040401356,
+      "eval_recall": 0.8348790689216221,
+      "eval_runtime": 1.7958,
+      "eval_samples_per_second": 222.185,
+      "eval_steps_per_second": 27.843,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 4.127802848815918,
+      "grad_norm": 4.313467025756836,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.217,
+      "loss": 0.2293,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8395201930584144,
-      "eval_loss": 0.2773597240447998,
-      "eval_precision": 0.8562091503267973,
-      "eval_recall": 0.8270594653573378,
-      "eval_runtime": 5.0754,
-      "eval_samples_per_second": 78.615,
-      "eval_steps_per_second": 9.851,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8518472677764712,
+      "eval_loss": 0.29089975357055664,
+      "eval_precision": 0.8599810186649794,
+      "eval_recall": 0.844880887434079,
+      "eval_runtime": 1.7949,
+      "eval_samples_per_second": 222.299,
+      "eval_steps_per_second": 27.857,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 2.408611297607422,
+      "grad_norm": 2.5997090339660645,
       "learning_rate": 2.5e-05,
-      "loss": 0.1966,
+      "loss": 0.2091,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8696741854636592,
-      "eval_f1": 0.844327731092437,
-      "eval_loss": 0.2846048176288605,
-      "eval_precision": 0.8411320530352577,
-      "eval_recall": 0.8477905073649754,
-      "eval_runtime": 5.0388,
-      "eval_samples_per_second": 79.185,
-      "eval_steps_per_second": 9.923,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8589543987905864,
+      "eval_loss": 0.29276683926582336,
+      "eval_precision": 0.8564068100358423,
+      "eval_recall": 0.8616566648481543,
+      "eval_runtime": 1.7944,
+      "eval_samples_per_second": 222.357,
+      "eval_steps_per_second": 27.864,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 2.3639743328094482,
+      "grad_norm": 3.3290369510650635,
       "learning_rate": 2.25e-05,
-      "loss": 0.199,
+      "loss": 0.198,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8545433391506689,
-      "eval_loss": 0.2909916341304779,
-      "eval_precision": 0.8639270714012982,
-      "eval_recall": 0.846653937079469,
-      "eval_runtime": 5.0638,
-      "eval_samples_per_second": 78.794,
-      "eval_steps_per_second": 9.874,
+      "eval_accuracy": 0.8796992481203008,
+      "eval_f1": 0.8569892473118279,
+      "eval_loss": 0.28468698263168335,
+      "eval_precision": 0.8522004241781549,
+      "eval_recall": 0.8623840698308783,
+      "eval_runtime": 1.8002,
+      "eval_samples_per_second": 221.645,
+      "eval_steps_per_second": 27.775,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 3.4640684127807617,
+      "grad_norm": 5.568458557128906,
       "learning_rate": 2e-05,
-      "loss": 0.187,
+      "loss": 0.1906,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8771929824561403,
-      "eval_f1": 0.846679266293906,
-      "eval_loss": 0.2870851755142212,
-      "eval_precision": 0.8609191655801824,
-      "eval_recall": 0.8356064739043463,
-      "eval_runtime": 5.0487,
-      "eval_samples_per_second": 79.031,
-      "eval_steps_per_second": 9.904,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.843111041207927,
+      "eval_loss": 0.31200090050697327,
+      "eval_precision": 0.8585673051692468,
+      "eval_recall": 0.831332969630842,
+      "eval_runtime": 1.7936,
+      "eval_samples_per_second": 222.458,
+      "eval_steps_per_second": 27.877,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 2.2104856967926025,
+      "grad_norm": 1.5475130081176758,
       "learning_rate": 1.75e-05,
-      "loss": 0.1812,
+      "loss": 0.1818,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8526315789473684,
-      "eval_loss": 0.2812826931476593,
-      "eval_precision": 0.8585304054054055,
-      "eval_recall": 0.8473813420621932,
-      "eval_runtime": 5.0484,
-      "eval_samples_per_second": 79.034,
-      "eval_steps_per_second": 9.904,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.850729517396184,
+      "eval_loss": 0.2906310558319092,
+      "eval_precision": 0.8535087719298247,
+      "eval_recall": 0.8481087470449173,
+      "eval_runtime": 1.795,
+      "eval_samples_per_second": 222.279,
+      "eval_steps_per_second": 27.854,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 3.044590950012207,
+      "grad_norm": 5.56436014175415,
       "learning_rate": 1.5e-05,
-      "loss": 0.1633,
+      "loss": 0.1756,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8596342841745197,
-      "eval_loss": 0.2956988215446472,
-      "eval_precision": 0.8555364857667042,
-      "eval_recall": 0.8641571194762684,
-      "eval_runtime": 5.0481,
-      "eval_samples_per_second": 79.04,
-      "eval_steps_per_second": 9.905,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.8514869535493182,
+      "eval_loss": 0.28101998567581177,
+      "eval_precision": 0.8523821128305106,
+      "eval_recall": 0.8506092016730314,
+      "eval_runtime": 1.7948,
+      "eval_samples_per_second": 222.311,
+      "eval_steps_per_second": 27.859,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.9169403314590454,
+      "grad_norm": 1.7677656412124634,
       "learning_rate": 1.25e-05,
-      "loss": 0.1607,
+      "loss": 0.174,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8696722245432793,
-      "eval_loss": 0.28752827644348145,
-      "eval_precision": 0.8706135006701596,
-      "eval_recall": 0.8687488634297145,
-      "eval_runtime": 5.0457,
-      "eval_samples_per_second": 79.078,
-      "eval_steps_per_second": 9.91,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8595070422535211,
+      "eval_loss": 0.2828814685344696,
+      "eval_precision": 0.8633733523114054,
+      "eval_recall": 0.8559283506092017,
+      "eval_runtime": 1.7987,
+      "eval_samples_per_second": 221.825,
+      "eval_steps_per_second": 27.798,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 1.9245001077651978,
+      "grad_norm": 2.1199889183044434,
       "learning_rate": 1e-05,
-      "loss": 0.1584,
+      "loss": 0.1705,
       "step": 1952
     },
     {
       "epoch": 16.0,
       "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8560793854229822,
-      "eval_loss": 0.2859440743923187,
-      "eval_precision": 0.8609538327526132,
-      "eval_recall": 0.8516548463356974,
-      "eval_runtime": 5.086,
-      "eval_samples_per_second": 78.451,
-      "eval_steps_per_second": 9.831,
+      "eval_f1": 0.855319904024935,
+      "eval_loss": 0.292202889919281,
+      "eval_precision": 0.862378106322743,
+      "eval_recall": 0.8491543917075832,
+      "eval_runtime": 1.7972,
+      "eval_samples_per_second": 222.011,
+      "eval_steps_per_second": 27.821,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 3.737988233566284,
+      "grad_norm": 2.521127223968506,
       "learning_rate": 7.5e-06,
-      "loss": 0.1535,
+      "loss": 0.1509,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8609292598654301,
-      "eval_loss": 0.29243767261505127,
-      "eval_precision": 0.8609292598654301,
-      "eval_recall": 0.8609292598654301,
-      "eval_runtime": 5.0479,
-      "eval_samples_per_second": 79.042,
-      "eval_steps_per_second": 9.905,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8568221901555235,
+      "eval_loss": 0.2991296947002411,
+      "eval_precision": 0.8596491228070176,
+      "eval_recall": 0.8541553009638116,
+      "eval_runtime": 1.7965,
+      "eval_samples_per_second": 222.104,
+      "eval_steps_per_second": 27.833,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 3.7625505924224854,
+      "grad_norm": 3.4539577960968018,
       "learning_rate": 5e-06,
-      "loss": 0.1432,
+      "loss": 0.1549,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8616171059774413,
-      "eval_loss": 0.29657039046287537,
-      "eval_precision": 0.859873949579832,
-      "eval_recall": 0.8634297144935443,
-      "eval_runtime": 5.0616,
-      "eval_samples_per_second": 78.829,
-      "eval_steps_per_second": 9.878,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.855319904024935,
+      "eval_loss": 0.3000391125679016,
+      "eval_precision": 0.862378106322743,
+      "eval_recall": 0.8491543917075832,
+      "eval_runtime": 1.7991,
+      "eval_samples_per_second": 221.772,
+      "eval_steps_per_second": 27.791,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.5611210465431213,
+      "grad_norm": 1.3508776426315308,
       "learning_rate": 2.5e-06,
-      "loss": 0.1466,
+      "loss": 0.1469,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8568221901555235,
-      "eval_loss": 0.29467570781707764,
-      "eval_precision": 0.8596491228070176,
-      "eval_recall": 0.8541553009638116,
-      "eval_runtime": 5.0458,
-      "eval_samples_per_second": 79.075,
-      "eval_steps_per_second": 9.909,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8609292598654301,
+      "eval_loss": 0.2942558228969574,
+      "eval_precision": 0.8609292598654301,
+      "eval_recall": 0.8609292598654301,
+      "eval_runtime": 1.8067,
+      "eval_samples_per_second": 220.841,
+      "eval_steps_per_second": 27.674,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 7.14449405670166,
+      "grad_norm": 3.1654489040374756,
       "learning_rate": 0.0,
-      "loss": 0.1411,
+      "loss": 0.1493,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8609292598654301,
-      "eval_loss": 0.2950780689716339,
-      "eval_precision": 0.8609292598654301,
-      "eval_recall": 0.8609292598654301,
-      "eval_runtime": 5.0549,
-      "eval_samples_per_second": 78.933,
-      "eval_steps_per_second": 9.891,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.855319904024935,
+      "eval_loss": 0.3026413023471832,
+      "eval_precision": 0.862378106322743,
+      "eval_recall": 0.8491543917075832,
+      "eval_runtime": 1.7968,
+      "eval_samples_per_second": 222.067,
+      "eval_steps_per_second": 27.828,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8460375995160000.0,
-      "train_loss": 0.2363556080177182,
-      "train_runtime": 1909.167,
-      "train_samples_per_second": 38.184,
-      "train_steps_per_second": 1.278
+      "train_loss": 0.24652244458433056,
+      "train_runtime": 623.7969,
+      "train_samples_per_second": 116.865,
+      "train_steps_per_second": 3.912
     }
   ],
   "logging_steps": 500,